The unprecedented boom in monolithic AI data centers is causing countries to tear up long-standing energy strategies and regroup around modular nuclear reactors to meet demand sustainably. It is also generating growth around the edge for real-time AI inferencing, which some telcos have moved to exploit. This in turn is causing some of them to consider how best to power these edge data centers cost-effectively, in accordance with whatever environmental targets they have. Yet one argument for more edge computing is that it saves power and will slow down the rampant growth in demand by centralized data centers. The mantra is to compute centrally when you must, and at the edge when you can. Verizon was peddling this line when…